WordPress, clearly. Of 527 US dental practice websites we scanned in June 2026, 54.8% run on WordPress, 35.5% run on a custom or unidentified system, 5.3% run on Wix and 4.2% on Squarespace. WordPress is the most common platform in six of the seven metros we scanned. Columbus, Ohio is the one exception.
A website's publishing platform decides more than how it looks. It decides how easily structured data gets added, how a theme handles page weight, and how much a practice owner can change without paying a developer. We read each site's HTML for the signatures WordPress, Wix, Squarespace and Shopify each leave behind, and classified everything else as custom or unidentified. This page covers what we found: the overall mix, how it shifts by metro, and one pattern in the data that runs against the usual assumption about which platform is "more set up" for AI search.
What platform do most dental websites use?
WordPress runs 289 of the 527 sites we identified, 54.8%. A custom or unidentified system is next at 35.5%, then Wix at 5.3% and Squarespace at 4.2%. One practice in our sample was classified as Shopify; the platform label alone does not establish how the site was used.
| Platform | Practices | Share of 527 |
|---|---|---|
| WordPress | 289 | 54.8% |
| Custom or unidentified | 187 | 35.5% |
| Wix | 28 | 5.3% |
| Squarespace | 22 | 4.2% |
| Shopify | 1 | 0.2% |
The July 2026 general-web comparison is not a matched one. W3Techs reports WordPress usage in two ways: among websites with a known content management system, and among all websites it tracks. Our dental percentages include the custom or unidentified category in their denominator. A known-CMS share therefore is not a matched comparison with our dental table; the populations and detection methods differ too. Use W3Techs for general-web context, not to infer that dentists adopt WordPress more or less often than other businesses.
Source: W3Techs, "Usage statistics of WordPress", checked 26 July 2026.
Does the platform mix change by metro?
Mostly it holds steady, with one exception. WordPress is the leading platform in six of the seven metros we scanned, running from 50.8% of practices in Nashville to 61.9% in Tampa. Columbus, Ohio breaks the pattern: 46.4% of the practices we read there run on a custom or unidentified system, against 42.9% on WordPress, the only metro in our sample where WordPress isn't in front.
| Metro | Practices | WordPress | Custom/unidentified | Leading platform |
|---|---|---|---|---|
| Charlotte | 153 | 59.5% | 32.7% | WordPress |
| Austin | 112 | 54.5% | 33.0% | WordPress |
| Raleigh | 64 | 56.2% | 34.4% | WordPress |
| Nashville | 59 | 50.8% | 37.3% | WordPress |
| Columbus | 56 | 42.9% | 46.4% | Custom/unidentified |
| Tampa | 42 | 61.9% | 31.0% | WordPress |
| Salt Lake City | 20 | 55.0% | 35.0% | WordPress |
We don't know why Columbus is different, and we're not going to guess at a reason the data can't support. It could be a strong local web-design shop that favors custom builds, a franchise pattern, or simple sample noise at n = 56. It's the kind of local anomaly a metro-level view catches and a national average would hide.
Source: Citevio dental practice web anatomy study, June 2026. Full metro breakdown and CSV: citevio.com/data.
The platform counts have a permanent citation in the archive under DOI 10.5281/zenodo.23143214.
Does the platform affect whether a site has structured data?
In our sample, recorded schema status differs by platform. WordPress sites are missing detected LocalBusiness or Dentist schema 64.4% of the time, the highest rate among the platform groups in this table. Squarespace sites are missing it 13.6% of the time, and Wix sites 28.6%. This does not show that the platform caused the difference.
| Platform | n | No detected LocalBusiness/Dentist schema | Both LocalBusiness & FAQPage |
|---|---|---|---|
| Squarespace | 22 | 13.6% | 0.0% |
| Wix | 28 | 28.6% | 7.1% |
| Custom or unidentified | 187 | 46.0% | 2.7% |
| WordPress | 289 | 64.4% | 3.8% |
We measured an association, not a mechanism. We did not check which themes, plugins or vendor settings produced the markup, so the table cannot explain the gap or establish what a platform adds by default. LocalBusiness is a description of a physical business or branch; Google documents it for details such as business hours and departments. Running WordPress alone does not establish whether that markup is present or correct.
Source: Citevio dental practice web anatomy study, June 2026, n = 527. Schema definitions: schema.org/LocalBusiness. What it's for: Google Search Central, "Local Business (LocalBusiness, Restaurant, Store) structured data", checked 26 July 2026.
Is WordPress better or worse for AI visibility?
We can't answer that from this dataset, and we're not going to pretend otherwise. Only 3 of 527 practices in our sample were visible in ChatGPT and 18 in Perplexity, far too few to split across five platform categories without the result becoming noise, not a finding.
The recorded visibility counts do not support a platform ranking. The raw records also include failed API calls; a failed call is not evidence that a practice was absent from an answer. The schema table measures detected markup, not whether an AI engine read it or used it to recommend a practice. Google says no special schema.org markup is needed for AI Overviews or AI Mode, and any structured data should match visible text. That guidance concerns Google Search's AI features; this dataset does not establish a schema requirement for ChatGPT or Perplexity.
Source: Google Search Central, "AI features and your website". Technical eligibility is separate from a practice being recommended.
If you'd rather see your own platform's numbers than average ones, the checker below reads your site directly.
Curious how fast your platform loads? See our companion study, how slow are dental websites.
What if your site is on a dental vendor platform?
Then this study does not have a number for you, and it is better to say so than to guess. The dental-specific website vendors many practices use were not tagged as their own bucket when Citevio scanned these 527 sites. A vendor-built site could match a WordPress or other detected signature, or fall inside the "custom or unidentified" category, which is 187 of 527 sites (35.5%). That category is not a count of dental vendor sites. We cannot tell you how those sites performed as a group, because we never separated them.
That means one sentence is off limits here, and it is the sentence a page like this is expected to produce: vendor platforms are not better or worse in our data, because our data does not contain the comparison. Anyone publishing a ranking of dental website vendors on AI visibility should be asked how they identified each one and how many sites were in each cell.
What does not depend on the vendor is the thing you can check yourself this afternoon. A platform is a way of producing HTML; what matters is the HTML it produces and what your server does with a request for it.
- Is your content in the page source? View the raw source of a treatment page and search it for your own service copy. If it is absent, investigate whether JavaScript supplies it later or whether the server returned a different page. Absence from one response alone does not identify the cause.
- Can you reach your own
robots.txt? Request the public file and inspect the response. If you cannot edit its rules, ask the vendor who can; inability to edit the file is different from inability to read it. - Can you add or edit a structured data block? Some vendor platforms allow it, some allow it only through a support ticket, and some not at all.
- Who controls the CDN or firewall? On a hosted platform this is usually the vendor, not you — which means the answer to "why is a crawler being refused" may live in their support queue rather than in your settings.
Those four are worth putting to your vendor in one email, and the answers say more about your position than any platform league table would. How to run the underlying checks yourself is on how to test if AI can read your dental website. If page speed is the reason you are asking, the companion study is how slow are dental websites. The first check on that list, content missing from the page source, is the exact failure mode covered in how JavaScript can hide a dental site from AI crawlers.
Do you need a new website, or three fixes to the one you have?
These measurements alone are not a reason to rebuild. They identify missing detected local-business markup and crawler refusals, but we did not test whether each platform lets an owner correct them in place. Ask who can change the page markup and access rules before deciding whether a platform switch is necessary.
The numbers behind that are narrow on purpose. In the June 2026 study of 527 practice sites, 53.9% had no detected LocalBusiness or Dentist schema. In a separate July 2026 test of 499 dental sites whose robots.txt allows AI crawlers, 72 (14.4%) were refused by their own server anyway, and in 12.6% of that sample Googlebot was served normally in the same run. These readings identify what to investigate; they do not establish the cause or whether your vendor gives you the controls needed to change it.
Sources: Citevio dental practice web anatomy study, 6–26 June 2026, n = 527. Citevio live crawler access study, 22 July 2026, n = 499 sites whose robots.txt allows AI crawlers and whose homepage loaded for a plain browser request. That second study is one day's measurement — hosting and CDN rules change without notice — and it measures access, not citation: removing a refusal was never measured to produce a mention. Which layer caused the refusal (host, security product, CDN, plugin or a human decision) was not recorded, and it is a different instrument from the robots.txt file study of 6,497 domains, so the two rates are not interchangeable.
This is not an argument that nobody should rebuild. Design, speed, accessibility, or simply being unable to edit your own pages are all real reasons to replace a site, and none of them was measured here. The decision should follow inspection of the specific obstacles we measured and the changes your vendor can actually make, rather than the platform label alone.
If you are handing this to a developer, the useful form is a request list rather than instructions. Ask them to: check what the server returns to AI crawler user-agents and record the status codes; make sure the page content is present in the raw HTML rather than assembled in the browser; and make the structured data match the text a visitor can see. Those are outcomes to verify, not a plugin to install — a specific plugin recommendation from us would be an untested one.
One question this page cannot answer: whether a separate site for the cosmetic side of a practice helps. We have not measured that, and our sample has no cosmetic-versus-general breakdown to answer it from, so there is no honest figure to give either way.
The refusal pattern is written up in full, with the method, on robots.txt open but AI bots blocked, and the tests you would run before deciding anything are on how to test if AI can read your dental website.
How we identified each platform
The sample is 527 US dental practice websites: 506 in Charlotte, Austin, Raleigh, Nashville, Columbus, Tampa and Salt Lake City, and 21 in 9 smaller nearby towns, scanned between 6 and 26 June 2026. For each site we read the page's HTML for a fixed set of signatures: wp-content or wp-json for WordPress, wix.com or _wix for Wix, the word squarespace for Squarespace, and cdn.shopify for Shopify. A site without any of those signatures is classified as "custom or unidentified," which covers hand-coded sites, page builders we don't specifically detect, and older systems alike, not a single platform of its own.
- A signature match, not a full audit. A site could carry a leftover signature from a migration, or hide its real platform behind a heavily customized theme. We did not manually verify every classification.
- "Custom/unidentified" is a catch-all. At 35.5% of the sample, it is large enough that a future pass naming its largest sub-groups, such as Joomla, Drupal, or specific page builders, would add real information.
- Sample, not census. This is 7 metros plus 9 nearby towns, not a national sample. We publish a national figure only once we can read at least 2,000 sites; this study does not meet that bar.
- A snapshot. Every number here carries the June 2026 scan window. A practice can migrate platforms at any time.
The full framework behind our scans, scores and city reports is in how we measure AI visibility, and the underlying dataset is free to download at citevio.com/data. If you think a number on this page is wrong, email contact@citevio.com with the domain and what you saw. We would rather be corrected than quoted wrongly.
Citevio's dental market research is published openly under CC-BY-4.0 with a DOI, so the numbers can be checked and reused; client data is never published.
Common questions
What platform do most dental websites use?
In our June 2026 study of 527 US dental practice websites, 54.8% run on WordPress, 35.5% run on a custom or unidentified system, 5.3% run on Wix and 4.2% on Squarespace. WordPress leads in six of the seven metros we scanned.
Is WordPress better than Wix or Squarespace for AI visibility?
We can't answer that from this dataset. Only 3 of 527 practices in our sample were recorded as visible in ChatGPT and 18 in Perplexity, too few to support a platform ranking; failed API calls are missing observations, not evidence of absence. WordPress sites in our sample more often lacked detected local-business markup than Wix or Squarespace sites. That is a different measurement from AI visibility, and does not establish that missing markup caused fewer recommendations.
Why is "Custom or unidentified" such a large share?
Our detector looks for specific signatures in a page's HTML: wp-content or wp-json for WordPress, wix.com for Wix, the word squarespace for Squarespace, and cdn.shopify for Shopify. Any site without one of those signatures, including hand-coded sites, older systems like Joomla or Drupal, and platforms we don't specifically detect, falls into "Custom/Unknown." It is a catch-all, not a single platform.
Does the platform mix change by metro?
Yes, in one clear case. WordPress is the most common platform in six of the seven metros we scanned. Columbus, Ohio is the exception: 46.4% of the practices we read there run on a custom or unidentified system against 42.9% on WordPress, the only metro where WordPress isn't in front.
How does WordPress's share among dental practices compare to its share of the general web?
The July 2026 general-web comparison is not a matched one. W3Techs distinguishes WordPress's share among websites with a known CMS from its share among all tracked websites. Our dental sample's 54.8% includes custom or unidentified sites in the denominator and uses a different population and detector. It does not establish whether dentists adopt WordPress more or less often than other businesses.
Why is Columbus, Ohio the one metro where WordPress isn't the leading platform?
We don't know, and we're not going to guess at a reason the data can't support. In Columbus, 46.4% of the 56 practices we read run on a custom or unidentified system against 42.9% on WordPress, the only metro in our sample where WordPress isn't in front. It could be a local web-design shop that favors custom builds, a franchise pattern, or simple sample noise at that size.
Does Shopify appear among dental practice websites in this study?
Exactly one of the 527 practices we scanned in June 2026 was classified as Shopify, 0.2% of the sample. The platform label alone does not establish whether the site was used as a storefront. The count is too small to draw a platform comparison; it is listed for completeness.
Could the platform detector have misclassified any sites?
Yes, and that's a stated limit of the method, not a hidden one. We read each site's HTML for a fixed set of signatures, wp-content or wp-json for WordPress, wix.com for Wix, and so on, and a site could carry a leftover signature from a past migration or hide its real platform behind a heavily customized theme. We did not manually verify every classification.
Is this platform breakdown a national statistic for all US dental practices?
No. This is a sample of 527 sites across 7 metros plus 9 smaller nearby towns, not a national census. We publish a national figure only once we can read at least 2,000 sites, and this study doesn't meet that bar yet.
What if my dental website runs on a dental-specific website vendor, not one of the platforms in this table?
This study does not have a vendor-specific number. Dental website vendors were not tagged separately when we scanned these 527 sites. A vendor-built site could match a WordPress or other detected signature, or fall into the custom or unidentified group of 187 sites, 35.5%. That group is not a count of vendor sites, and we cannot compare vendors because we never separated them out.
Do I need to switch off WordPress, or any platform, to fix these problems?
These measurements alone do not justify a platform switch. Missing detected local-business markup and crawler refusals are issues to investigate, but we did not test whether every platform allows an owner to correct them in place. Ask the vendor who can change the markup and access rules. Design, speed and editing limits are separate reasons to consider a rebuild, not conclusions from this platform comparison.
Could this platform mix have changed since the June 2026 scan?
Yes. Every number on this page carries that scan window, taken between 6 and 26 June 2026, and a practice can migrate platforms at any time afterward. Treat the percentages as a snapshot of that period, not a live count of what's running today.
Does Citevio recommend one platform over another for a new dental practice website?
No, and this data isn't built to support that kind of recommendation. We can't say whether WordPress, Wix or Squarespace performs better for AI visibility, because only 3 of 527 practices in our sample were visible in ChatGPT and 18 in Perplexity, far too few to split across platforms without the result becoming noise. What we can point to is the pattern we did measure with a large enough sample: which platforms in our data are more or less likely to have detected local-business markup in place.
How reliable is the metro-by-metro breakdown, given some metros have far fewer practices than others?
Take the smaller metros with more caution. Salt Lake City's row is built from 20 practices and Nashville's from 59, against 153 in Charlotte and 112 in Austin. The table displays the metros that meet the dataset's 20-practice privacy threshold; smaller towns are included in the overall count but have no row here. A few category changes would shift the smallest metros' percentages more than Charlotte's.